Coquin-Viennot Danièle, Moreau Stéphanie
Laboratoire Langage, Mémoire, et Développement Cognitif (LMDC), University of Poitiers, France.
Br J Educ Psychol. 2007 Mar;77(Pt 1):69-80. doi: 10.1348/000709905X79121.
Understanding and solving problems involves different levels of representation. On the one hand, there are logico-mathematical representations, or problem models (PMs), which contain information such as 'the size of the flock changed from 31 sheep to 42' while, on the other hand, there are more qualitative representations, or qualitative situation models (QSMs), comprising events expressed in everyday terms, such as 'a lot of lambs had been born'.
We set out to show that an apparent contradiction between a PM and a QSM can result in poorer performances, especially when the activation of formal representations is not fully automatic.
We tested 44 third years (mean age 8;6) and 46 fourth years (mean age 9; 8) from two primary schools in France.
Change problems and compare problems were drafted in two versions: one where the QSM and PM were consistent and one where they appeared to be contradictory but were not (referred to hereafter as the inconsistent version). For example, if the numerical data proved that the size of a flock of sheep had increased (PM), the consistent version would state that a lot of lambs had been born (QSM), whereas the inconsistent version would state that the wolf had devoured some of the sheep (QSM). Each pupil was given 7 consistent problems and 7 'inconsistent' ones to solve.
For the change problems, errors were more frequent on the inconsistent versions, especially among the younger subjects. For the compare problems, there were more errors on the inconsistent versions regardless of the subjects' age.
It is in situations where the problem schema is not automatically activated (younger subjects or compare problems) that the QSM plays its role of intermediate representation and causes performances to deteriorate if it does not correspond to the PM.
理解和解决问题涉及不同层次的表征。一方面,存在逻辑 - 数学表征,即问题模型(PMs),其中包含诸如“羊群数量从31只羊变为42只”之类的信息;另一方面,存在更具定性的表征,即定性情境模型(QSMs),由日常用语表达的事件组成,例如“出生了很多小羊”。
我们试图表明,问题模型(PM)和定性情境模型(QSM)之间明显的矛盾会导致表现较差,特别是当形式表征的激活不完全自动时。
我们测试了来自法国两所小学的44名三年级学生(平均年龄8岁6个月)和46名四年级学生(平均年龄9岁8个月)。
变化问题和比较问题被设计成两个版本:一个版本中QSM和PM是一致的,另一个版本中它们看似矛盾但实际上并非如此(以下简称不一致版本)。例如,如果数字数据证明一群羊的数量增加了(PM),一致版本会说出生了很多小羊(QSM),而不一致版本会说狼吃掉了一些羊(QSM)。每个学生要解决7个一致问题和7个“不一致”问题。
对于变化问题,不一致版本的错误更频繁,尤其是在较年轻的受试者中。对于比较问题,无论受试者年龄如何,不一致版本的错误都更多。
在问题模式未自动激活的情况下(较年轻的受试者或比较问题),定性情境模型(QSM)起到中间表征的作用,如果它与问题模型(PM)不对应,就会导致表现变差。